Data Management Plans:Policy and Practice

Workshop

CASRAI Reconnect 2014 Conference

Ernie Boyko & Wendy Watkins, Carleton UniversityChuck Humphrey, University of Alberta

Outline➢Workshop objectives➢Context for Data Management Plans (DMPs)

○Why research data management○Tri-Council+ initiative○Value propositions and uptake

➢An exercise➢DMP purpose, process, and practice

○How to produce a DMP

Overall objectives

➢ To provide you with practical advice in preparing Data Management Plans using online DMP tools.

➢ The background and the context for DMPs➢ A tool that you can use (here or at home)

Data management planning

Data management planning consists of all the activities related to data in the data lifecycle.

Managing research data ➢ The process of controlling the research administrative

information and data generated during a research project.

➢ An integral part of the research process○ DM is particularly challenging when studies involve

several researchers and/or when studies are conducted from multiple locations.

➢ Depends on ○ how they are produced and used throughout the

research lifecycle,○ the types of data involved,○ how data are collected, stored, shared, & preserved.

(Adapted from Penn State University Libraries Publishing and Curation Services: http://www.libraries.psu.edu/psul/pubcur/what_is_dm.html)

Benefits of research data management planning

➢ Helps to identify issues and strategies early in your research project.

➢ Confirms that you have documented your compliance with institutional and funder policies and ethics approval requirements.

➢ Provides focus on data sharing and re-use opportunities of your research data.

➢ Charts a pathway for your research data to remain useful and to be preserved for future use.

➢ Ensures that any questions regarding your findings can be validated using your data.

(Adapted from the University of Sheffield, University Library http://www.sheffield.ac.uk/library/rdm/dmp)

DMPs: form and function

➢A data management plan is a formal statement that describes how a researcher will address specific topics on the production and stewardship of data both during and after a research project is completed.

➢ It facilitates research data management to support the research project as well as broader institutional and funder objectives.

➢Sustain data sharing and re-use.

How data management planning relates to DMPs

➢ Provides context for DMPs➢ Specific policy aspects related to data

management get expressed in DMPs➢ Describes data management in research

administrative concepts➢ Connects data management with larger

goals and objectives of funders and research administrators

Some context for research data management in Canada

➢ This is not a new activity in many leading research countries.

➢ Researchers in Canada are doing this in an informal or ad hoc manner now but without the benefit of a systematic process.

➢ Canada’s Tri-Councils+ are about to launch an initiative that will bring this to the fore.

➢ And as we do so,Canada will be catching up to other regions of the world.

Tri-Agency / CFI Consultation – Fall 2013

•“Capitalizing on Big Data: Toward a Policy Framework for Advancing Digital Scholarship in Canada”

•Discussed ways the agencies can promote excellence in data management practices and help to establish a “Culture of Stewardship” in Canada

•Proposed that data management plans be required as part of grant applications

•Highlighted the need for ongoing coordination with other players in the « research ecosystem »

Source: Christine Trauttmansdorff, Executive Director, Corporate Strategy and Performance, SSHRC) Presented to Canadian Association of Research Libraries, Québec – November 2014

GoC Open Government / Open Science

•Government of Canada initiative to maximize access to federally-funded scientific research, to encourage greater collaboration and engagement with the scientific community, the private sector, and the public.

•G8 Science Ministers Agreement: "to promote policies that increase access to the results of publicly funded research results to spur scientific discovery, enable better international collaboration and coordination of research, enhance the engagement of society and help support economic prosperity".

Source: Christine Trauttmansdorff, Executive Director, Corporate Strategy and Performance, SSHRC) Presented to Canadian Association of Research Libraries, Québec – November 2014

Source: Christine Trauttmansdorff, Executive Director, Corporate Strategy and Performance, SSHRC) Presented to Canadian Association of Research Libraries, Québec – November 2014

Tri-Agency Approach: Key Considerations

•Data management requirements will impact researchers, students and institutions, among others

•Long-term success depends on availability of standards, repositories, expertise and other resources.

•Education and awareness activities, sharing best practices, etc. are essential

•Coordination, consultations and communications will be required on an ongoing basis

•Parallel initiatives being led by the GoC (Treasury Board Secretariat) on Open Government and Open Science

Source: Christine Trauttmansdorff, Executive Director, Corporate Strategy and Performance, SSHRC) Presented to Canadian Association of Research Libraries, Québec – November 2014

Source: Christine Trauttmansdorff, Executive Director, Corporate Strategy and Performance, SSHRC) Presented to Canadian Association of Research Libraries, Québec – November 2014

Tri-Agency: Next StepsFall/winter 2014-15:•Tri-agency working group with RDC and CAURA observers

•Participation in discussions such as CSPC, CASRAI, CAURA, CARL, etc.

•Environmental scan: Kathleen Shearer

•External advisory group

•Promotion and awareness

2015:•Public consultation on proposed policy direction

•Guiding principles, pilots, guidance documents, etc.

•Determine next steps

Source: Christine Trauttmansdorff, Executive Director, Corporate Strategy and Performance, SSHRC) Presented to Canadian Association of Research Libraries, Québec – November 2014

Value propositions for DMPs

➢Data sharing➢Reproducible research➢Managing digital assets➢Building a corpus of Canadian research

data

Data sharing➢Data archives have been enabling the sharing of

data for over 5 decades (e.g.,ICPSR, UKDA, Carleton)

➢2004 OECD▪ Declaration on Access to Research Data from Public Funding

➢UK Research Councils▪ “Publicly funded research data are a public good, produced in

the public interest, which should be made openly available with as few restrictions as possible in a timely and responsible manner that does not harm intellectual property.”

➢US National Science Foundation▪ “[A Data Management Plan (DMP)] should describe how the

proposal will conform to NSF policy on the dissemination and sharing of research results.”

Reproducible Research

➢Supports academic integrity➢Journals increasingly asking for data to be

deposited as condition of publishing➢Compliance with funders’ expectations

that research is reproducible

Management of digital assets

➢Digital content (data, etc) must be managed to facilitate access and use over the long run

➢Data stewardship“acquire, manage, organize, preserve, and provide access to massive amounts of data for use and re-use by a variety ofinterdisciplinary and heterogeneous communities over time”http://besser.tsoa.nyu.edu/howard/Talks/imagem-som.pdf

A Canadian corpus of research data

➢Creates opportunities to create new data from existing data.

➢Allows the application of new methodologies to existing data.

➢Facilitates connections between research outputs and the data associated with them to deepen the knowledge behind the findings.

Excercise

➢ What is the value proposition with respect to Data Management Plans for your institution or situation?

➢ Take some time to reflect on this and then at your table, discuss the value propositions with your colleagues.

Examples of DMP uptake

➢International Polar Year DMP (2006)➢Genome Canada (2009)➢U.S. National Science Foundation (2011)➢Research Councils of the U.K. (RCUK) (2010-12)➢U.S. Department of Energy (July 2014)➢International Development Research Centre

DMPs in practice➢ Development of templates and tools

○ The central themes in an organization’s data policy act as the topics of a DMP template. Within these topics, relevant questions are asked about each area. A researcher then provides her or his answers to each question.

○ Tools to help with a DMP■ DMPOnline (UK Digital Curation Centre)

● DMP Builder (Canada U of Alberta)■ DMP Tool (US California Digital Library)

➢ Establish a consultation process to support the use of DMPs

A DMP consultation process

Meet with a librarian and review the

DMP process

Start a new research project

Prepare an initial draft of a

DMP

Third-party review of the

draft DMP

Incorporate reviewer’s

suggestions in the DMP

Submit DMP

DMP purposes

➢ A DMP embodies the main elements of an organization’s data policy.

➢ The following example is based on the University of Alberta Research Policy and its focus on the stewardship of research records .

University of Alberta Research Policy

University of Alberta Research Policy

“Ensure that principles of stewardship are applied to research records, protecting the integrity of the assets.”

“[DM] plans recognize the stages through which research records will be produced, managed, documented, stored, disseminated and deposited (with either a staging or a preservation repository). Furthermore, these plans will identify the data stewards across a project’s lifecycle.”

DMP elements for digital assets

➢Funder and data description➢Data collection➢Documentation and metadata➢Storage and backup➢Data sharing and re-use➢Preservation➢Responsibilities and resources

(This template is taken from the University of Alberta’s Data Management Plan Builder, https://dmp.library.ualberta.ca/projects/my-project-university-of-alberta-template--130?show_form=yes )

Open Discussion

DMP elements for long-term access

➢Sources for S&T data collection➢Data collection storage and backup➢Documentation and metadata➢Choice of data repository➢Terms of data deposit➢Terms of data access➢Responsibilities and resources

Knowledge Mobilization Plans➢ Scholarly publication lifecycle➢ Not unlike DMPs and data lifecycle

Exercise

➢Prepare a DMP using DMP Builder (shortened template) around depositing data in a publication with a repository.

Hypothetical Example

➢Hypothetical open access and data policies for the Canadian Science Policy Centre

➢Open Access: all research publications of the CSPC will be freely available to the public through the CSPC website.

Hypothetical Example

➢Data Policy: all data used as evidence in research publications of the CSPC will be made available to allow the findings to be reproduced.

○A citation for each dataset in an article or report will be included.

○All data will be discoverable through the use of persistent identifiers (handles or digital object identifiers) and preferably registered with DataCite Canada.

Hypothetical Example

➢Data Policy: all data used as evidence in research publications of the CSPC will be made available to allow the findings to be reproduced.

○Intelligent access to data will be provided to ensure that confidentiality is protected.

Intelligent Access

Intelligent access employs data management methods and services that best align user needs with the access control deemed necessary for a particular dataset.

Examples of intelligent access:• using visualization techniques to explore data without having

to transport it, • co-locating “big data” with high performance computing, • aggregating data to guard against disclosure while

safeguarding sensitive content, • providing secure environments for analyzing sensitive data,

and• performing data linkages in a controlled setting.

Hypothetical Example

1. Will this CSPC publication involve the use of data?a. No.

i. No DMP requiredb. Yes.

i. Sources of the data1. Data collected by the researcher2. Data from other data providers

ii. Requirement for special software or code1. No: go to next item2. Yes: steps to provide access to code

Hypothetical Example

2. Are there any restrictions on the data being used?a. No.

i. Go to the next item.b. Yes.

i. Confidentiality of research participants1. Need for informed consent about reuse of the

dataii. Data from another producer

1. Does the producer allow the data to be redisseminated?

Hypothetical Example

3. Will you deposit your data with a data repository? a. No.

i. How will you make your data available?b. Yes.

i. Which data repository will you use?1. Does this repository issue persistent identifiers,

such as DOIs, for research data?a. No.

i. How will you obtain a persistent identifier for your data?

Hypothetical Example

3. i. Which data repository will you use?1. Does this repository issue persistent identifiers for research

data?a. No.

i. How will you obtain a persistent identifier for your data

b. Yes.i. Will this data repository register data DOIs

with DataCite?1. No.

a. Manually enter DOI2. Yes.

a. Requirement met

Hypothetical Example

3. i. Which data repository will you use?2. Does this repository support intelligent access?

a. No.i. If 2b-i is No, exitii. If 2b-i is Yes, how will you protect

confidential information?b. Yes.

i. If 2b-i is No, exitii. If 2b-i is Yes, select an appropriate way to

protect confidential information

Hypothetical Example

4. Data citationa. Has a citation format for datasets been identified?